Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Need to optimize SE-ResNeXt 152 model on ImageNet dataset on Fluid. #8624

Closed
qingqing01 opened this issue Feb 28, 2018 · 1 comment
Closed

Need to optimize SE-ResNeXt 152 model on ImageNet dataset on Fluid. #8624

qingqing01 opened this issue Feb 28, 2018 · 1 comment
Projects
Computer Vision on Fluid

Comments

@qingqing01
Copy link
Contributor

qingqing01 commented Feb 28, 2018 •
edited
Loading

以下数据来自 @BigFishMaster , Fluid和PyTorch的对比：

paddle-P40-1-card(无内存优化), max batch size=10, speed(fp+bp): 1.00sec
paddle-P40-1-card(有内存优化), max batch size=25, speed(fp+bp): 1.90sec
paddle-P40-8-card(无内存优化), max batch size=80, speed(fp+bp): 2.35sec
pytorch-P40-1-card, max batch size=48, speed(fp+bp): 2.06sec
pytorch-P40-8-card, max batch size=384, speed(fp+bp): 2.77sec
@qingqing01 qingqing01 mentioned this issue Feb 28, 2018
Closed
@jacquesqiao jacquesqiao mentioned this issue Feb 28, 2018
Closed
@BigFishMaster
Copy link
Contributor

add:

pytorch-P40-1-card, max batch size=48, speed(fp+bp): 2.06sec
pytorch-P40-4-card, max batch size=192, speed(fp+bp): 2.50sec
pytorch-P40-8-card, max batch size=384, speed(fp+bp): 2.77sec

@qingqing01 qingqing01 moved this from TODO to DONE in Computer Vision on Fluid Apr 13, 2018
@qingqing01 qingqing01 moved this from DONE to DOING in Computer Vision on Fluid Apr 13, 2018
@qingqing01 qingqing01 moved this from DOING to DONE in Computer Vision on Fluid May 12, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
Computer Vision on Fluid
  
DONE
Milestone
No milestone
Development

No branches or pull requests
2 participants
@BigFishMaster @qingqing01